基于自适应八叉树的点云数据压缩方法研究  

Adaptive Octree-Based Data Compression Method for Points Cloud Data

作　　者：程效军[1,2] 李伟英[1] 张小虎[1] 

机构地区：[1]同济大学测量与国土信息工程系,上海200092 [2]现代工程测量国家测绘局重点实验室,上海200092

出　　处：《河南科学》2010年第10期1300-1304,共5页Henan Science

基　　金：国家科技支撑计划课题(2006BAJ03A07)

摘　　要：首先详细讨论了借助包围盒建立点云K邻域以及使用平面拟合方法获取法矢量等方法,然后根据点云数据法矢量变化程度,采用自适应八叉树得到压缩后的点云数据.对相关参数的选取以及算法步骤进行了改进.最后,使用此方法实现了点数为10000的点云模型的数据压缩.Firstly, the establishment of K-nearest points by bounding box as well as calculation of the normal vectors by plane fitting was discussed in detail in this paper. Then,according to the change ratio of normal vector, octree-based compression method was proposed to obtain the compressed data. Some key parameters and algorithm were also improved in this paper. Finally, point cloud data which consist of 10 000 points was compressed by the methods.

关 键 词：点云压缩 K邻域 法向量拟合 八叉树 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]